Chinese semiconductor industry

Status
Not open for further replies.

european_guy

Junior Member
Registered Member
What metric is used to compare who wins/loses? Overall revenue? Revenue from foundry business? Technical capability? Or just plain wafer capacity? Or profit margin?

I think the technical capability + capacity indices gap between tsmc and competition will actually grow bigger in the next few years. tsmc’s expansion plan in the most advanced nodes is higher than its closest competition. tsmc has not peaked yet.

I was thinking about Technical capability / Technology level.

I agree until 2024 / 2025 gap will get wider, but then Intel will start to close it.

On very long term speculation, say 10-15 years, I think the geopolitical considerations are more important than the technical ones.

The key point is that both US and China strongly want to become technological leaders in semiconductor. This has been stated by both leadership, China leadership even gave a target of 2035 for "realizing socialist modernization".

On the US part, well, we now how US operates, they will use any mean to reach the target. If Intel is able to close the gap with TSMC on their legs alone by 2026, well that's great, if not they will force TSMC to give it to Intel, for instance forcing a "licensing" of TSMC technology to Intel, or forcing Intel entering as shareholder in TSMC. The key point is that US government owns Taiwan government and so TSMC can't really prevent Intel getting the technology if this is what the US government wants.

Regarding China, well, in a much more elegant way, also China, as a superpower, has the possibility and the resources to close the gap once EUV technology will be available. For instance see the developing of SMIC 7nm, many sources report that it is very similar to TSMC 7nm, and this makes sense because SMIC chief scientist is from TSMC. In addition to this, SMIC will get the full and unlimited support of the whole Chinese semiconductor industry togheter with state research institutes and big funds: from Chinese Academy of Sciences (CAS) to AMEC and NAURA, from top universities to the last supplier, from Chinese IC manufactures customers to State Big Fund. This is a much more powerful display of forces than what TSMC can ever dream.

So in both cases, in the long term, is an uneven battle for TSMC. And TSMC cannot win it.
 

ansy1968

Brigadier
Registered Member
I was thinking about Technical capability / Technology level.

I agree until 2024 / 2025 gap will get wider, but then Intel will start to close it.

On very long term speculation, say 10-15 years, I think the geopolitical considerations are more important than the technical ones.

The key point is that both US and China strongly want to become technological leaders in semiconductor. This has been stated by both leadership, China leadership even gave a target of 2035 for "realizing socialist modernization".

On the US part, well, we now how US operates, they will use any mean to reach the target. If Intel is able to close the gap with TSMC on their legs alone by 2026, well that's great, if not they will force TSMC to give it to Intel, for instance forcing a "licensing" of TSMC technology to Intel, or forcing Intel entering as shareholder in TSMC. The key point is that US government owns Taiwan government and so TSMC can't really prevent Intel getting the technology if this is what the US government wants.

Regarding China, well, in a much more elegant way, also China, as a superpower, has the possibility and the resources to close the gap once EUV technology will be available. For instance see the developing of SMIC 7nm, many sources report that it is very similar to TSMC 7nm, and this makes sense because SMIC chief scientist is from TSMC. In addition to this, SMIC will get the full and unlimited support of the whole Chinese semiconductor industry togheter with state research institutes and big funds: from Chinese Academy of Sciences (CAS) to AMEC and NAURA, from top universities to the last supplier, from Chinese IC manufactures customers to State Big Fund. This is a much more powerful display of forces than what TSMC can ever dream.

So in both cases, in the long term, is an uneven battle for TSMC. And TSMC cannot win it.
@european_guy fully agree bro and here I hope , Huahong should up its game and Huawei stay motivated. Due to escalating cost and SMIC possible success, Huawei may leave the FAB business and concentrate on its core. Their FAB attempt is a short term fix until SMIC can guarantee them the supply of chips.
 

mrandolph

New Member
Registered Member
I was thinking about Technical capability / Technology level.

I agree until 2024 / 2025 gap will get wider, but then Intel will start to close it.

On very long term speculation, say 10-15 years, I think the geopolitical considerations are more important than the technical ones.

The key point is that both US and China strongly want to become technological leaders in semiconductor. This has been stated by both leadership, China leadership even gave a target of 2035 for "realizing socialist modernization".

On the US part, well, we now how US operates, they will use any mean to reach the target. If Intel is able to close the gap with TSMC on their legs alone by 2026, well that's great, if not they will force TSMC to give it to Intel, for instance forcing a "licensing" of TSMC technology to Intel, or forcing Intel entering as shareholder in TSMC. The key point is that US government owns Taiwan government and so TSMC can't really prevent Intel getting the technology if this is what the US government wants.

Regarding China, well, in a much more elegant way, also China, as a superpower, has the possibility and the resources to close the gap once EUV technology will be available. For instance see the developing of SMIC 7nm, many sources report that it is very similar to TSMC 7nm, and this makes sense because SMIC chief scientist is from TSMC. In addition to this, SMIC will get the full and unlimited support of the whole Chinese semiconductor industry togheter with state research institutes and big funds: from Chinese Academy of Sciences (CAS) to AMEC and NAURA, from top universities to the last supplier, from Chinese IC manufactures customers to State Big Fund. This is a much more powerful display of forces than what TSMC can ever dream.

So in both cases, in the long term, is an uneven battle for TSMC. And TSMC cannot win it.
As it is now it may seem far-fetched. But can the political winds in Taiwan change? Is there a way to avoid giving away the crown jewels to the US, given the political will?

I think we are going into a semiconductor slump with low demand and oversupply. In this situation Intel would be under even more pressure and the US politicians may lose interest, certainly if it became obvious that this is about bailing out losers.

And maybe it finally dawns on them that a rocket that blows up shit doesn't really require a 2nm chip.
 

ansy1968

Brigadier
Registered Member
As it is now it may seem far-fetched. But can the political winds in Taiwan change? Is there a way to avoid giving away the crown jewels to the US, given the political will?

I think we are going into a semiconductor slump with low demand and oversupply. In this situation Intel would be under even more pressure and the US politicians may lose interest, certainly if it became obvious that this is about bailing out losers.

And maybe it finally dawns on them that a rocket that blows up shit doesn't really require a 2nm chip.
@mrandolph Bro here's a way out, a JV or create a Chinese subsidiary, use domestic equipment and concentrate on the Chinese market, right now TSMC can leverage their tech cause its only rival Samsung is having trouble with their 7nm, 5nm and maybe their 3nm. BUT China doesn't have an EUVL, that's why TSMC is being blackmailed by the US. IF China can provide an alternative I'll bet many will jump ship making the US restriction useless, sadly now is not the time and waiting until 2025 is not an option.
 

Weaasel

Senior Member
Registered Member
My my, have I been at least a week or probably even more behind, but browsing through the Web I today learned that SMIC has succeeded in producing 7 nm chips that replicate or at least are very similar and comparable to those of TSMC, TSMC using DUV lithography equipment already in its possession. The various sources that I have read state that regardless of the fact that the yields that SMIC can obtain to make these 7 nm chips will be significantly less than if it were in possession of EUV machines and even if the chips are actually not as advanced as TSMC 7 nm chips, they are still much more advanced that any chips that SMIC or any other know Chinese company has been known to produce commercially and are likely more advanced than the ones that Intel can produce.

It all makes sense now... It all makes sense that in recent months the United States has greatly stepped up the urgency of the pressure and demands that it has made to the Netherlands and Japan to prevent them from selling advanced DUV equipment and other IC chips manufacturing equipment, materials, and components used in manufacturing IC chips. It is very likely that the United States got wind of SMIC's progress and certainly it knew of its efforts and had assessed that there was a great likelihood that with equipment, components, and materials already in its possession and that it has access to, whether domestically sourced or foreign, that SMIC would be able to make substantial progress in making highly advanced IC chips even to the extent of producing the most advanced chips that are possible using DUV.

I have read from this forum and other sources that SMIC and other Chinese companies and entities have stockpiled much in terms of imported DUV equipment, components, and substances used in producing the most advanced DUV made chips to the extent that China is almost sanctions proof with regards to DUV. Combined with China's own domestic efforts at producing 28 nm or finer DUV equipment, as well as China already being able to produce high quality photoresists and also having companies that produce equipment for vapour deposition, etching, and ion implantation that are as good as those anywhere else, China has done well to keep the US in check and has significantly reduced its vulnerabilities to US IC chip and semiconductor technology blackmail.

It all makes sense as to why China's response to US threats were just like a calm shrugging silence...

One can only expect the United States now to do its utmost to ensure that China does not externally obtain the technology to produce the most advanced qualitatively and in terms of high yield to produce even more advanced chips than the DUV made 7 nm TSMC replicas. But China can however breath a lot more now than was the case a few to several years ago...
 

ansy1968

Brigadier
Registered Member
My my, have I been at least a week or probably even more behind, but browsing through the Web I today learned that SMIC has succeeded in producing 7 nm chips that replicate or at least are very similar and comparable to those of TSMC, TSMC using DUV lithography equipment already in its possession. The various sources that I have read state that regardless of the fact that the yields that SMIC can obtain to make these 7 nm chips will be significantly less than if it were in possession of EUV machines and even if the chips are actually not as advanced as TSMC 7 nm chips, they are still much more advanced that any chips that SMIC or any other know Chinese company has been known to produce commercially and are likely more advanced than the ones that Intel can produce.
@Weaasel bro that's the crux of it all, they can't accept the fact that even with the restriction, SMIC still beat Intel with one hand tie on its back ..lol
It all makes sense now... It all makes sense that in recent months the United States has greatly stepped up the urgency of the pressure and demands that it has made to the Netherlands and Japan to prevent them from selling advanced DUV equipment and other IC chips manufacturing equipment, materials, and components used in manufacturing IC chips. It is very likely that the United States got wind of SMIC's progress and certainly it knew of its efforts and had assessed that there was a great likelihood that with equipment, components, and materials already in its possession and that it has access to, whether domestically sourced or foreign, that SMIC would be able to make substantial progress in making highly advanced IC chips even to the extent of producing the most advanced chips that are possible using DUV.
Bro ASML had stood its ground , let's applaud them for doing so.
I have read from this forum and other sources that SMIC and other Chinese companies and entities have stockpiled much in terms of imported DUV equipment, components, and substances used in producing the most advanced DUV made chips to the extent that China is almost sanctions proof with regards to DUV. Combined with China's own domestic efforts at producing 28 nm or finer DUV equipment, as well as China already being able to produce high quality photoresists and also having companies that produce equipment for vapour deposition, etching, and ion implantation that are as good as those anywhere else, China has done well to keep the US in check and has significantly reduced its vulnerabilities to US IC chip and semiconductor technology blackmail.

It all makes sense as to why China's response to US threats were just like a calm shrugging silence...

One can only expect the United States now to do its utmost to ensure that China does not externally obtain the technology to produce the most advanced qualitatively and in terms of high yield to produce even more advanced chips than the DUV made 7 nm TSMC replicas. But China can however breath a lot more now than was the case a few to several years ago.
That's the clincher for me that China may have something up their sleeves.
 

hvpc

Junior Member
Registered Member
The writer(s) of that article are hoping that their conclusions and assumptions are correct...
Technically, the authors are correct. But they are not looking at the whole picture as they are talking about the past.

UMC produced 14nm bitcoin mining chips as far back as 2017. Yet no one give them credit for having true 14nm capability. So, no one is treating SMIC any differently on their accomplishment last year with the MinerVA chip.

SMIC has since last year been improving on their capability and is now in risk production of real 7nm chips now.

So, the article is correct in stating SMIC could not build a real 7nm chip last year. But the article is incorrect in not realizing SMIC is now capable of a 7nm chip as of today. The yield may not be the best, but it's good enough to declare they could make 7nm chips.
 
Status
Not open for further replies.
Top